MRM-DIFF tutorial

Edited in 2014/10/06

Introduction

MRM-DIFF is a data processing tool for multiple reaction monitoring (MRM)-based

differential analysis. The main target of this application is ‘lipidomics’. The MRM transition, i.e.

precursor-product m/z pair, can be theoretically determined from in silico MS/MS database such as

LIPID MAPS and LipidBlast. This program hunts every peaks detected by each MRM transition by

means of correlation optimized warping (COW) based non-linear alignment. In addition, pooled QC

(quality control) data sets will be helpful for the automatic reference file picking and the peak

detection method (see manuscript). The features of MRM-DIFF are:

1. Every peaks detected by each transition are utilized

2. The identification and quantification results can be manually curated by the graphical user

interface (GUI).

3. It supports all data processing steps and pooled QC data sets would be helpful its processes.

Actually, the compound identifications should be performed by means of three or four

transitions for the determination of lipid class and fatty acid compositions. Such targeted analysis can

be performed by our reported program, MRMPROBS (see MRMPROBS section).

MRM-DIFF has been developed as the collaborative work among RIKEN, Osaka University,

and Reifycs Incorporation.

Hiroshi Tsugawa

RIKEN Center for Sustainable Resource Science

hiroshi.tsugawa@riken.jp

MRM-DIFF screenshot

Table of Contents

Software environments .............................................................................................................................. 3

Required software programs and files ...................................................................................................... 4

Downloading the ABF converter from Reifycs Inc. ............................................................................... 5

File conversion ......................................................................................................................................... 6

Reference library for MRMPORBS program (tab-delimited text format) .......................................... 7

Starting MRM-DIFF ................................................................................................................................... 9

Starting up your project ....................................................................................................................... 10

Importing Abf files ................................................................................................................................ 11

Setting parameters ............................................................................................................................... 12

MRM-DIFF viewer .................................................................................................................................... 14

Mouse operation in the chromatogram viewer ................................................................................... 14

Tool button ............................................................................................................................................. 15

Tab .......................................................................................................................................................... 16

Button .................................................................................................................................................... 17

List Box .................................................................................................................................................. 18

Details on the MRMPROBS function ...................................................................................................... 19

File menu ............................................................................................................................................... 19

Data processing menu........................................................................................................................... 20

Statistical analysis menu ..................................................................................................................... 21

Normalization setting ....................................................................................................................... 21

Statistical analysis setting................................................................................................................ 24

Identification menu ............................................................................................................................... 25

Option menu .......................................................................................................................................... 26

Export menu .......................................................................................................................................... 27

Appendix A: how to obtain appropriate file conversion of the Shimadzu .lcd file. .............................. 28

Suitable method file (.lcm) ....................................................................................................................... 28

Appendix B: mzML file conversion via ProteoWizard. ........................................................................... 32

Software environments

Microsoft Windows XP, -Vista, -7 or -8

.NET Framework 4.0 or later

Required software programs and files

Reifycs Analysis Base File Converter (ABF file converter)

Download link: http://www.reifycs.com/english/AbfConverter/

MRM-DIFF

Download link: http://prime.psc.riken.jp/Metabolomics_Software/MRM-DIFF/index.html

Reference library (tab-delimited text file)

Example: http://prime.psc.riken.jp/Metabolomics_Software/MRM-DIFF/index.html

Demonstration files

Download link: http://prime.psc.riken.jp/Metabolomics_Software/MRM-DIFF/index.html

MRM-DIFF can import Analysis Base File (ABF) format data. This program extracts

chromatogram data together with the reference library including the name of the target metabolite, its

retention-time and amplitude information, and precursor m/z and product m/z. The supported formats

for ABF conversion are Shimadzu Inc. (.LCD), Agilent Technologies (.D), AB Sciex (.WIFF), and

Thermo Fisher Scientific (.RAW).

MRM-DIFF is also acceptable to mzML format file converted by an open source file translator

ProteoWizard. Although our abf converter doesn’t accept Waters (.RAW) file due to the license

problem yet, MRM-DIFF can import Waters files via mzML.

MRM-DIFF program is implemented as a part of MRMPROBS software. Therefore, please

select ‘MRM-DIFF project’ in the new project window.

Downloading the ABF converter from Reifycs Inc.

1. Go to http://www.reifycs.com/english/AbfConverter/.

2. Check the requirements and license terms, and download the converter.

File conversion

1. Start “AnalysisBaseFileConverter.exe”.

2. Drag & drop MS vendor files into this program.

3. Click “Convert”.

4. The ABF files are generated in the same directory as the raw data files.

Reference library for MRMPORBS program (tab-delimited text format)

Four items are required in the library file in tab-delimited format. The first header’s name is

flexible but the item order should be followed.

1 column. Compound name

2 column. Precursor m/z (accurate m/z information is rounded into nominal m/z information)

3 column. Product m/z

4 column. Retention time [min]

Notes 1: The compound name should be entered in English one-byte characters.

Note 2: Sometimes the tab-delimited file exported from Microsoft Excel includes unexpected hidden

trailing columns. These unexpected columns after the ‘Retention time’ column cannot be handled by

MRM-DIFF. You can inspect the exported file by selecting a few rows (see below). If there are selected

characters after the last column (Retention time), edit the file in Excel to delete these columns and re-

export it again.

Good example (no unexpected column)

Bad example (there are unexpected columns)

Starting MRM-DIFF

Note that again, MRM-DIFF is run as a part of MRMPROBS program. Therefore, the assembly

name, i.e. EXE file name, is ‘MRMPROBS.exe’.

1. Starting up your project

2. Importing Abf files

3. Setting parameters

4. Running the software (1-2 min / sample)

*The tutorial uses 37 demonstration files and the lipid reference library which are downloadable

from the above link. The common measurement conditions of the demonstration files were as

follows.

Liquid chromatography: total 45 min run per sample with InertSustain C18: 2.1×150 mm, 3 m (GL

sciences Co.).

Mass spectrometer: MRM method with positive and negative ion mode.

Target metabolite number: 284

Total transitions: 284

The detail of experimental conditions is downloadable at the MRM Database section (ODS-lipids).

http://prime.psc.riken.jp/Metabolomics_Software/MrmDatabase/index.html

Starting up your project

1. File New project.

2. Chose a project type (select the bottom one for this demonstration).

Importing Abf files

Note:

We recommend that the project folder be made for each batch experiment. In the MRM-

DIFF project, three folders (raw, processed, aligned) and one file (*.mth) are generated.

They should be included in the same directory.

The file name should be entered in half-width alphanumeric symbols.

Select the file type of each file from “Sample”, “Standard”, and QC”. QCs must be

required for MRM-DILL program to perform chromatogram alignment- and peak

detection methods as well as LOESS-based normalization method. If you don’t have

pooled QC data sets, use the wild type (control) sample data sets as QCs. What matters

is to select the (biological or technical) replicate data sets. In such case, the LOESS-

Cubic spline normalization will not work well, but the others such as chromatogram

alignments and peak detections will be fine.

Decide the class ID used for the color labels.

The analytical order and class ID can be changed after data processing.

Setting parameters

Select ‘20140919_MRMDIFF_ReferenceLibrary_40mLC_SerumProject.txt’ and set the above

parameters for this demonstration.

[Recommended]

Peak detection

Smoothing method: linear weighted moving average.

Smoothing level: 1-2

Minimum peak width: 3-5

Minimum peak height: 50-100

Peak identification

Retention time tolerance: As long as the reverse phase or hydrophilic interaction

chromatography LC are used, 0.1-0.2 min is recommended.

Amplitude tolerance: non-meaningful

Minimum posterior: 50-70.

Advance: MRMDIFF

Column type: select a type used in your experiment.

Segment size: add the average peak width in your data sets.

Others: Min slack, max slack, and border limit are automatically determined from the column

type. These settings are derived from our pre-experiment. For the classical correlation optimized

warping (COW), select ‘Constant’ for the border limit and in such case, max slack parameter is

not meaningful.

MRM-DIFF viewer

Mouse operation in the chromatogram viewer

View mode

①. Chromatogram window: drag holding left click chromatogram scroll, drag holding right click

chromatogram zoom.

②. Detected window: left double-click the reverse triangle change the focused peak

③. Retention time window: drag holding right click warping on retention time range.

④. Intensity window: drag holding right click warping on intensity range.

Edit mode

①. Right click and drag on un-detected peak area detect new peak. Right click and drag on

detected peak area delete detected peaks.

②. Left click and drag on the peak edge [red square] change the location of the peak edge.

Tool button

File: start new project, open existing project, save as a project, and save the project.

Data processing: for data re-processing per file, per metabolite, or in all data sets.

Statistical analysis: data normalization and statistical analysis.

Window: non-meaningful in MRM-DIFF project.

View: change focused chromatograms.

Identification: Manual curation for identification results.

Option: re-define class ID and analytical order, choose the internal standard, decide “include” or

“exclude” data for statistical analysis.

Export: The result is exported in tab-delimited text format.

Help: show version information.

Tab

Chromatogram: All data manipulation tasks are performed here.

Raw data matrix: Not used in MRM-DIFF project.

Processed data matrix: Not used in MRM-DIFF project.

Statistical result: The result of statistical analysis is shown here.

Button

Reset: Reset the display range of chromatograms.

View: If you push this “View” button, the chromatogram viewer is changed to “Edit” mode. In

the “Edit” mode you can modify the peak edge and detect new peaks manually

None: The properties of detected peaks are shown in this ComboBox. You can confirm isotopic

ions and identified lipids.

Height: You can set the quantification mode. The default is set by peak height. Instead, you can

change it to area mode. By using the “All” option, the quantification mode is reflected =

implemented in all files and all metabolites.

Aligned: You can see raw chromatograms as well as just smoothed chromatograms.

List Box

If you double-click a transition name, the chromatograms are generated in the chromatogram viewer.

If you double-click a file name, the reference chromatogram used for alignments and peak detections

is highlighted.

Details on the MRMPROBS function

File menu

New project: used for creating a new project.

Open project: used for opening an existing project. Make sure that *.mth file, raw folder, and

processed folder are included in the same directory.

Save as: use to save as a new file.

Save: use to overwrite an existing project.

Data processing menu

Data re-processing can be done by newly optimized parameters in this option. Re-

processing is also performed per transition. Also, in MRM-DIFF program, you can re-set the

compound library and the identifications can be only done by checking ‘Re-identification processing

only’.

Statistical analysis menu

Normalization setting

At first, you can set properties of aligned peaks and files. In the file properties (left), you can

reset file type, class ID, or analytical order. If you clear the check box of the “Included” column, the

corresponding data are no longer used in the statistical analysis. In the alignment properties (right),

you can set internal standard information for each aligned peak. Please make sure to assign

“Annotated peak name” in the “internal standard” column.

Then, in this demonstration, please choose ‘Yes’ from the below message.

Select ‘20140919_MRMDIFF_FormulaLibrary_40mLC_SerumProject.txt’ as shown in below. The

compound names, which should be the same as the identified name in the MRM-DIFF program, and

formulas are utilized to estimate the peak abundance from isotopic ion. Moreover, the MRM-DIFF

program can also estimate the isotopic abundances from unknown peaks as ‘alkane’. Checking ‘Also

consider unknown peaks’ is to consider it.

Finally, select a normalization approach.

Note:

None: after implementation of the missing value approach, the values of the raw data matrix are

stored in the processed data matrix.

Internal standard: after implementation of the missing value approach, the value divided by the

internal standard value set in the “Option menu” is stored in the processed data matrix.

LOESS: after implementation of the missing value approach, the signal intensities of each

metabolite are normalized with the QC samples information by means of loess/cubic spline.

Internal standard + LOESS: After internal standard normalization, loess/cubic spline based

normalization is performed.

Statistical analysis setting

You can do principal component analysis. Add the calculated number of the principal components

and choose the scale and transform method.

Identification menu

You can manually correct identification result. This option may be useful to check internal

standards which are not included in the reference library.

Option menu

You can set properties of files. You can reset file type, class ID, or analytical order. If you

clear the check box of the “Included” column, the corresponding data are no longer used in the

statistical analysis.

Export menu

A tab-delimited text file can be exported for a raw data matrix, a processed data matrix, the

updated library, detected peak information detail, and PCA results. Moreover, the PCA result can be

exported by some image formats.

Appendix A: how to obtain appropriate file conversion of the Shimadzu .lcd file.

Suitable method file (.lcm)

Although you can do a content change of the .lcd file after LC-QqQ/MS (MRM) analysis, it is very

useful to construct a suitable method file (.lcm format file) for the successful file convert of the

MRMPROBS software.

1. Event name and channel (MRM transitions) rule.

2. Update compound table

After the method construction of MRM transitions, you should update the compound table m/z by

the MRM event. If you can analyze the samples by using the updated method file, you do not have

to perform any other tasks for the stable file convert.

For stable convert of Reifycs file convert

software, the compound name should be

made just by ASCII format.

MRM transitions should be

constructed for one metabolite.

The completely same precursor and

product m/z pair cannot be

acceptable in the file converter.

You can check the updated table by Method->Data Processing Parameters->Compound tab.

3. If your data (.lcd) were not collected by a suitable method described above, you can improve

the .lcd file by using the method file modified in the above way. After the construction of the

modified method file, please open “Postrun Analysis” of LabSolutions.

After selecting the analysis files (.lcd) push the “Apply to Method” button.

Select the modified method file and improve your .lcd file including the compound table m/z. If you

can do this, the file (.lcd) is successfully converted by Reifycs Inc. software.

4．File convert

Conditions: You can convert from .lcd files to .abf files on your computer by installing LabSolutions

software. “TTFLDataExportVer5.dll” of LabSolutions ver. 5.53 SP4 or later is required for the file

convert. Check the “TTFLDataExportVer5.dll” (Program Files (or *86)>LabSolutions) file property.

If the file size is less than 577,536 bytes, contact Shimadzu Inc. for a file change.

After “AnalysisBaseFileConverter.exe” is opened, drag and drop the .lcd files to this converter.

Push the “Convert” button. The ABF format files will be generated in the same folder as the .lcd

files.

Reifycs Inc. software refers to this compound table for

the file convert from .lcd file to .abf file.

Appendix B: mzML file conversion via ProteoWizard.

Required software and file

MSConvert

Download link: http://proteowizard.sourceforge.net/

Download ProteoWizard

1. Select download type: Windows installer (includes vendor reader support) is recommended.

2. Read license agreements and download the proteowizard.

(http://proteowizard.sourceforge.net/downloads.shtml)

Setup ProteoWizard

1. Follow the wizard windows. (Maybe you don’t miss it.)

2. “SeeMS” should be also imported.

Convert the vendor’s MS file to mzML via ProteoWizard

1. Open the MSConvertGUI.exe.

2. Select “List of Files”.

3. Select the vendor’s file via “Browse” button.

4. In the “Options”, never check any additional compression including “Use numpress linear

compression”, “Use numpress short logged float compression”, and “Use numpress short positive

integer compression”. Each of binary encoding precision is available.

5. Click “Start” button.